My name is Amir Saffari and this is my website and blog. I’ve a PhD in Computer Vision and Machine Learning and work at Alexa AI, Amazon as a Principal ML Scientist. Currently, I focus on generative AI, training large language models (LLMs), teaching LLMs to use thousands of tools and APIs to accomplish personalised and complex tasks in real-time, reasoning using weak supervision, and reinforcement learning based program synthesis.

Amir

Recent Papers

Complete list of papers

News

  • 2022, Mintaka dataset is now available, see our paper for more information.
  • 2021, Our two new papers at EMNLP extend differentiable knowledge graphs for complex QA, E2EQA and Rigel.

Recent Posts

CVPR 2010 Papers

I together with other colleagues got three papers accepted at CVPR 2010. Amir Saffari, Martin Godec, Thomas Pock, Christian Leistner, Horst Bischof, “Online Multi-Class LPBoost“, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2010. Code. Bernhard Zeisl, Christian Leistner , Amir Saffari , Horst Bischof, “Online Semi-Supervised Multiple-Instance Boosting“, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2010. Jakob Santner, Christian Leistner, Amir Saffari, Thomas Pock, Horst Bischof, “PROST: Parallel Robust Online Simple Tracking“, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2010.
Read more →

Patched isotostick.sh for Jaunty Jackalope Ubuntu

Since udevinfo has been removed and merged at binary level with udevadm in Jaunty Jackalope (Ubuntu 9.04), the isotostick.sh script which copies the ISO image to a USB stick does not work out of box. So here is a patched version of isotostick which uses udevadm instead. Download: isotostick-jj.sh
Read more →